A CONTINUOUS-TIME FORMULATION OF NONLINEAR MODEL-PREDICTIVE CONTROL

A model predictive approach is used to derive a continuous-time, nonli near feedback control law for open-loop stable, single-input single-ou tput processes with deadtime. The derived model predictive control law minimizes a quadratic performance index in the presence of input cons traints and process deadtime. A key feature of the control law is that , for its implementation, one does not need to perform an online optim ization. In the absence of constraints, the control law in special cas es is identical to the feedback controllers that have already been der ived using a geometric approach. The connections between the derived c ontrol law and (a) the globally linearizing control and (b) the linear internal model control are established. The model predictive approach provides some insight into the problem of constraint handling in geom etric control methods. The application of the theory is illustrated by a chemical reactor example.